There are many clinical studies directly demonstrating the importance of hypoxia in limiting the response to essentially all forms of cancer therapy. Additionally, hypoxia is thought to affect the very nature of cancer, with hypoxic tumors being more susceptible to genetic instability and aggressive phenotypes. At present, it is not known whether hypoxia is simply associated with aggressive tumors, or whether it is the driving force. There has been a great interest in the development of methods to image clinically relevant tumor hypoxia by non-invasive means. In this Application we will employ a 2-nitroimidazole hypoxia marker, EF5, for this purpose. EF5 is unique because its oxygen-dependent bioreduction can be studied in two distinct ways: first;highly specific antibodies can be used for quantitative immunohistochemical detection at sub-cellular resolution (e.g. fluorescence microscopy of tissue sections or flow cytometric analysis) and secondly;non-invasive imaging using F-18-1abeled drug. Both methods have been shown to correlate with individual-tumor radiation-response prediction in the 9L gliosarcoma rat-tumor-model. Using this model, three hypotheses will be tested (a fourth will employ the HTIOB0 human sarcoma xenograft): The first aim will test the hypothesis that moderate to high concentrations of hypoxia markers are necessary to promote optimal prediction of tissue hypoxia. This represents a continuation of our prior grant and requires the engineering of an innovative method (developed by Dr. Olof Solin in Finland) to produce high specific activity fluorine gas (not possible with normal techniques). The technology developed in this aim will benefit planned human studies, and allow new methods for labeling compounds with radioactive fluorine. The second Aim will compare the now validated radiation response prediction of EF5 with two other clinically relevant techniques: uptake of labeled Cu-ATSM and interstitial fluid pressure. The goal in this aim is to assess whether the various predictive assays are truly monitoring hypoxia, or other relevant resistance factors. The 3rd Aim will test the clinically relevant hypothesis that F-18-EF5 imaging can detect intra-tumoral changes in radiation resistance caused by heterogeneity in the spatial distribution of hypoxia. Success of this Aim could lead to important new uses of therapies such as IMRT and proton beams which can be directed to spatial locations of one cc or so. Our final Aim, to be studied in the HT1080 human sarcoma xenograft, will test the hypothesis that F-18-EF5 imaging can monitor microenvironmental changes caused by antivascular and antiangiogenic therapies. In summary, the aims of this grant will extend and advance our overall laboratory interest in quantifying the degree and extent of clinically relevant hypoxia in the tumor microenvironment.